The Weird, Wild Future Of CGI

CGI has revolutionized film, art, and design over the past 35 years. Here’s what the next 35 might look like.

The Weird, Wild Future Of CGI

The Uncanny Valley was first defined by robotics professor Masahiro Mori in 1970. Mori noticed a very curious reaction in people who came in contact with humanoid robots. The humanoid robots that were clearly robots–think Bender–caused positive, normal reactions in people. But Mori observed that humanoid robots that look like humans but aren’t exactly human–think creepy current androids like this–provoke an uncanny feeling, generating negativity and even revulsion. However, the moment robots become indistinguishable from real human beings–like the robots in Westworld–the feeling disappears. Hence the uncanny “valley,” where robots sit on one side, and completely believable androids sit on the other side. The middle–the valley–is occupied by robots that almost look human.


The concept of the Uncanny Valley was later adapted to computer graphics to describe the unpleasant sensation of seeing something on the screen that looks photorealistic but doesn’t feel exactly right, because it doesn’t match up to our perception of the physical world, like the characters of the 2007 film Beowulf.

Will the Uncanny Valley exist in the future? How will the phenomenon evolve alongside technology? In his video essay Goodbye Uncanny Valley, the 3D video artist Alan Warburton takes a look into the current state and future of hyper-realistic computer graphics. While his essay is focused on film, it raises questions about what future media, design, and architectural rendering will look like in general–and makes me think of a near-future where people will be able to choose the alternate reality in which they want to live in.

The Future

Warburton begins his essay announcing that we’ve conquered this Uncanny Valley. It’s almost true. Most CGI in movies–natural landscapes, weather phenomena, buildings, and all kinds of objects–is now indistinguishable from reality. CGI humans, however, still elicit that uncanny feeling, just like real androids. For instance, take the Star Wars film Rogue One, which featured the most advanced 3D humans yet: The late actors Peter Cushing (as Grand Moff Tarkin) and Carrie Fisher (as Princess Leia). But even while their effort was amazing, jarred audiences immediately detected the fake humans. They looked extremely realistic, but they didn’t feel like real people. In fact, the effect was so eerie that it was a distraction.

Despite that quibble, the rest of Warburton’s analysis is solid. He explains his vision of the evolution of graphics using a map, divided into four different regions. There’s the familiar Uncanny Valley, but also the Frontier, the Beyond, and the Wilderness.

The Uncanny Valley, of course, is the period we all know, stretching from the mid-1980s (with movies like The Abyss, Terminator 2, or Jurassic Park) to the present (with the aforementioned Rogue One or Blade Runner 2049). Over the past 35 years, companies like ILM, Weta, Pixar, Framestore, and Blue Sky have developed new modeling, animation, and rendering technologies to convincingly simulate everything from rocks and water to fur and herds of animals. These technologies were sometimes created for specific films or by software companies. Others were the result of academic research published in papers and demoed at Siggraph–the always awesome annual conference for CGI nerds.


Eventually, all these technologies were included in commercial software packages that anyone could buy. The advances in hardware technology–especially the massive processing power of dedicated graphics cards–made it possible for anyone to have access to 3D rendering abilities that only a few years ago were exclusively reserved for production houses with huge render farms. Combined with mature off-the-shelf 3D software, this led to the commoditization of computer graphics. Now, anyone with talent, dedication, and a PC can create stuff at the same level as the best Hollywood blockbusters.

So when anyone can create anything, what comes next? Warburton calls this “the Frontier” of the industry. He argues that the only way big production houses can distinguish themselves is to go for the biggest, the best, and the most awesome. Stuff that small companies and individuals can’t do. And for that, you need scale: Hordes of 3D artists creating mind-blowing giant scenes at a rapid pace, exemplified by periodic sagas like the Marvel Universe and Star Wars. He forgets to point out another way in which production houses are using economies of scale. Movies that take place in the same fictional universes, or that share certain recurrent themes like space, allow them to reutilize old 3D elements.

Recently, an executive at one of the biggest VFX companies told me how crucial it is for them to build upon existing elements (like using parts from a spaceship in Movie A in another spaceship in Movie B) in order to remain competitive in a field with increasingly ridiculous profit margins. Warburton argues that this “Frontier” of CGI is not so much about innovation but about “commerce, construction, and consolidation,” a period in which studios only want to build “upwards and outwards, at greater speeds, volumes, and frequencies, with greater details and density.”

The Distant Future

After the Frontier lies what Warburton calls “the Beyond.” This is where things get really interesting. He believes the future of CGI may present a big problem: the post-truth phenomenon, or the paradoxical effect in which the more realistic images get, the less people believe they’re real.

You’ve probably heard people saying “Oh, that’s CGI!” even when it’s not. Warburton mentions the Tumblr Hyperrealcg, which features real photos falsely presented as renderings to make fun of this common mistake. The truth is that, while that site is a joke, we are quickly heading into a territory in which nobody will be able to distinguish reality captured with a camera from synthetic reality made at home with a personal computer. And when anyone can create media that can’t possibly be identified as synthetic, our current problems with fake viral news will seem insignificant in comparison. Imagine the use of fabricated videos for propaganda in wars or politics. Even worse: Think on a smaller scale, like students bullying other students by distributing false, embarrassing videos or angry lovers spreading lies about their exes. What will happen in court if forensic experts can’t say if something is real or not? Sure, invisible watermarks can prevent some of these abuses, but they aren’t foolproof. All of this could lead to a world in which reality can lose its value completely.

Warburton juxtaposes the post-truth phenomenon with something he calls “theoretical photorealism.” This is the use of CGI to illustrate a phenomenon that has never been observed, like when a physicist and a visual effects team used theoretical models and data to render a black hole in Christopher Nolan’s movie Interstellar. The paradox here is that science is using CGI to bring reality to things that, until now, were only visible through mathematical formulas–while elsewhere, people are calling out real imagery as CGI.


That brings us to Warburton’s final area of focus: what he calls “the Wilderness.” I’m a fan of the people living in the Wilderness. It’s a place where 3D artists embrace CGI’s current imperfections, instead of focusing on seamlessly merging it with reality. For these CGI artists, the only thing that matters is using 3D as a form of expression, not only disregarding its faults but making them into art. Both the artists and the public know that it’s all a farce; the fun is in embracing the artifacts that arise in the rendering process. “Work here exposes the magic trick,” Warburton says. The Wilderness is a “digital grotesque,” a place with a “wonderful lineage that encompasses pantomime, Shakesperian farce, political cartoons, punk, club culture, comedy, reality TV, and, most recently, politics.”

And so it is! You’ll find evidence of this burgeoning genre of CGI across the web: Warburton mentions the Wilderness equivalent of Pixar, Cool 3D World. Another example? The Instagram account Ifyouhigh, or the work of artists like Roger Kilimanjaro or Esteban Diacono:

/how’s the week going?/ #octane #cinema4d #houdini @juxtapozmag @hypebeast @designboom @plastikmagazine

A post shared by Esteban Diacono (@_estebandiacono) on

I think there’s a fifth zone, in addition to Warburton’s four proposed areas of development. With the rise of artificial intelligence, augmented reality, and virtual reality, we’re headed into what I would call “the Unknown.” In a few decades, the physical world we consider reality will dissolve into many computer-generated worlds with a billion dimensions, indistinguishable from reality. The borders of real and electronic life will be erased forever, in a way that books like Ready.Player.One can barely articulate. Machine-to-brain interfaces will allow us to sink or soar into these worlds. We’ll be able to make our own, just by thinking of them, the same way we dream at night. Like in the 17th century Spanish play Life Is a Dream, our reality will become whatever we choose to believe it is.

And for many, that won’t be the physical world we know today.

About the author

Jesus Diaz founded the new Sploid for Gawker Media after seven years working at Gizmodo, where he helmed the lost-in-a-bar iPhone 4 story. He's a creative director, screenwriter, and producer at The Magic Sauce and a contributing writer at Fast Company.